On Mon, Feb 17, 2014 at 3:36 PM, Andrew Puschak <apuschak@gmail.com> wrote:
>
> The only difference
> seems to be that they come from Access through the ODBC connector. We are
> only using one database in the cluster and everything is relatively simple,
> although high volume.
This would lead to a problem in the ODBC connector, even if I don't
understand what would make a single statement to become so slow. Any
chance you can run the same statement thru an ODBC connecter via a
test program?
Luca
